1. Field of the Invention
The present invention relates to a signal processing apparatus, a signal processing method, and an image pickup apparatus, and more particularly to a smear amount correction technique of correcting the smear phenomenon-relevant image degradation of an image sensor.
2. Description of the Related Art
Conventionally, a light receiving surface of an image sensor such as a CCD has been roughly divided into an optical black portion (hereinafter, referred to as “the OB portion”) 4001 for shielding light not to be incident on a pixel and an effective pixel portion 4002 for receiving light as shown in FIG. 5. One of the reasons is for providing the OB portion 4001 to correct a dark current noise by subtracting the output the signal values of the OB portion 4001, i.e., the output signal values generated without light incident from the output signal values of the effective pixel portion 4002.
It is known that an image sensor such as a CCD generates a noise referred to as “smear” when a high-luminance object is in an angle of view thereof. When intense light enters the effective pixel portion 4002 of the image sensor, light leaking into a vertical transfer unit of an image sensor (not shown) is converted into an electric signal to highlight a signal on a vertical line. As a result, an image is colored in magenta on the vertical line of the vertical transfer unit into which the light leaks, and further, if light is large in quantity, the image may be subjected to clipped highlight.
In the image sensor, the determination of a difference between the output signal value of the OB portion (specifically, an OB-value reference portion 4003) disposed on a lower portion of the effective pixel portion 4002 and the output signal value on the same vertical line allows a smear amount to be corrected. If successive images are subjected to correction of the smear amount, a cyclic filtering processing is sometimes performed which refers to the smear amount on the previous frame as shown in FIGS. 6A to 6C.
If the cyclic filtering processing shown in FIGS. 6A to 6C is performed, its cyclic coefficient is generally controlled in consideration of the motion of an image. For example, if the image does not move, the cyclic coefficient is increased to thereby use also information of the OB-value reference portion 4003 of the previous frame, which reduces the influence of noises on the vertical line to enable the smear amount to be effectively corrected. For example, in the motion of an image is detected, the position of smear is also moved, so that information of the OB-value reference portion 4003 of the previous frame cannot be used. In this case, setting the cyclic coefficient to a small value allows the overcorrection to be reduced.
In the conventional smear amount correction, however, a plurality of OB lines is processed by a fixed cyclic coefficient, so that the results of image motion detection increases the weighting for the specific OB line, thereby increasing the influence of the noise (refer to Japanese Laid-Open Patent Publication (Kokai) No. H07-67038, for example). For this reason, there has been proposed a method of correcting the smear amount which is insusceptible to noises from the specific OB line independently of the results of the image motion detection. For example, there has been proposed a method of making it difficult to be insusceptible to the noises from the specific OB line by calculating the cyclic coefficient such that the weightings of the cyclic coefficients for the plurality of OB lines of the image sensor can be equal to one another (refer to Japanese Laid-Open Patent Publication (Kokai) No. 2006-166368, for example).
In the smear amount correction described in Japanese Laid-Open Patent Publication (Kokai) No. 2006-166368, the cyclic coefficients are calculated with attention drawn only to the absolute value of an amount of change in the smear amount, so that the cyclic coefficient is calculated as is in the case where the smear amount increases, even when the smear amount suddenly decreases. Therefore, the conventional smear amount correction processing makes it difficult to effectively suppress decrease in image quality due to the overcorrection made when the smear amount suddenly decreases. However, the smear amount overcorrected more than the originally generated smear looks worse to the user, which requires to suppress the overcorrection.